Monday, May 31, 2010

Recoveries: Burnin' Dominos

Early 2009, I started a remodeling project of sorts.  I had a nice coopered vanity from Lowes', but it had a black granite top with a white vessel bowl on it. -yawn-.  My replacement was a pie-wedged top of ribbon Sapelé with an onyx vessel bowl.

I did the glue-up of the pie wedges with Dominos since those end-grain glue-ups would be horribly weak.  I used a lot of Dominos.  Thing is, I planned on pattern routing the new top against the old with an offset bearing to make it slightly larger. Personal matters intervened for a year and everything sat.

I forgot the part about the larger pattern.  My Domino placement required the offset.  mea culpa.

I don't have a picture of it, but I burned through not one, not two, but three Dominos.  You know what I mean: the profile was this lovely Sapelé then a gaping hole with white steamed Beech (the Domino).  Good grief.

When I hit the first one, I remembered the offset, but much too late now.  I continued to route exposing 3 total.

So... how to fix?

First, put down power tools, step away from pointy objects, and cuss like nobody's business. Won't fix it, but you'll relax :)

I don't have pictures of the burned Dominos since that was before I thought blogging would be cool (both of you readers should let me know if that's true...)  However, imagine the profile with a gash showing the Beech Domino.  Since the profile hit the Domino at a skew angle, the hole is oblong.  This was my fix (best to read it all as some skipped details will be immediately clear when you see where things were going):

Get an offcut of the project that has the same grain you are matching.  Matching ribbon Sapelé is no fun.  Find a part of the piece that if placed in the burned hole would be a perfect match.  Here, I circled the area in chalk.

Determine the size of the mortise to know which bit to use.  I had used 10mm Dominos so that was the cutter to use (I still remembered the fence setting used to mortise the holes... if you don't, you could use a larger cutter; it will be clearer in a moment).  Figure out which mortise width setting will equal or exceed the width of the revealed hole.  In my case, 2 were using the middle width setting, one used the widest setting.

Plunge a mortise directly over the burned hole.  The intention is to leave a clean Domino mortise with no remnants of the originally revealed hole. I have no photos of this.

Next, make a shallow plunge with the Domino into your scrap in the target area.  You just want to mark the board so plunge enough that the reciprocating motion cuts across the whole mortise, but not far enough to reduce the remaining width.  In this picture, I plunged further than I normally would just so it would show clearly in the photo.

Take the offcut to the bandsaw and carefully bandsaw a perfect Domino replacement by tracing the outside edges of the shallow mortise.  You're bandsawing a very small piece: plan you cuts appropriately to stay away from the blade.  In my case, I made a rectangle that encompassed the mark and sanded the corners.

You can see where this is going, right?  Put this now perfectly fitted matched face-grain Domino into the hole you mortised over the burned hole and glue in place.  Profile it again the next day.

Here's the result after I put a finish on it, which is honestly the best way to judge the grain match.  One of the burned Dominos was at the joint on the profile.  I'll admit one more trick... the burned hole to the left of the previous picture (one with a red clamp) looked great, but not perfect... the eye caught it mostly because there wasn't a joint line there.  Nothing a 0.5mm pencil couldn't fix and it completely went away.

Ultimate Wide-Base of the OF-1400

Everybody discovers this when they get their new Festool OF-1400 router, but I'll point it out to those who don't have it to figure it out.

Routing on an edge always leaves at least half the base off the stock making the router less stable.  Add to that the higher center of gravity of a plunger like the OF-1400 and you can easily make a mess of that edge.

The OF-1400 comes with two guide rods intended for use with the edge guide or the guide stops (the pads that let you run it on the rail).  Thing is, the guide stops sit level with the router so I use them all the time as a very extended handle and wide-base for the router.  Used this way, it is nearly impossible (oh, but I'll find a way...) to tip the router on the edge.

In this configuration, the pad is out nearly a foot from the bit!  The added stability is amazing and offers a lot of control for guiding the router.  On a recent project with a lot of a shaped profile requiring climb-cutting on material very prone to chip-out, I used this configuration to bump-route much of the edge; the little bit of climb-cutting I did was very safe and controlled as that long arm to the base gives you a lot of mechanical advantage.

Now, I like to use the edge guide even with a bearing-guided bit for certain situations where I want to make multiple passes rather than hog off the whole profile in one shot.  You can still get the advantage of this wide-base setup by simply extending the rods as far as possible to the opposite side of the router.

Saturday, May 29, 2010

MFT-1080, 'merican Sized

My MFT-1080 was always a lean-over affair, and that gets old.  Eventually, the MFT/3 was released.  It comes in 2-3 inches higher.  Perfect work height.

Well, I don't have an MFT/3 and couldn't in any way justify replacing the 1080 with one.  So I built a table that sits underneath the MFT-1080 bringing it up to the same height as the MFT/3.  This will be useful if the CMS ever makes it out here as it is designed to attach to the MFT/3 or stand beside it.

Voilà my table. Sorry, messy.  The table is just 4 legs, a table set down 3" from the top of the legs and a vanity panel in the back mostly used to give the table more rigidity.  There are stretchers front to back on the legs

The table sits on the top of the legs.  There are glue-blocks behind where the MFT legs sit that are beveled so the table falls into place easily, but the final location has no play.  Note that I took the legs off.  They are trivial to re-attach.  It gives me more clearance on the table (super handy for tools).  Besides, I don't take it out often.

My bench to the left and the miter saw stand to the right are all at the same height, which makes working with long stock much easier especially when cross-cutting on the MFT.  The gap between the bench and the MFT is partly due to the MFT's fence being in a position that would cover a door to a hidden clamp compartment.  At first this annoyed me, but it is super handy to have the trash can by the bench especially when using the jigsaw to rough-cut boards over the gap; chips fall into the trash.

Now, I like to use the benches for planing.  The front of the MFT is coplanar to the front of the table so I get extra support when I clamp a panel to the side for trimming.  Thing is, even the MFT-1080 on its own is wiggly when planing, even without hitting the mechanical resonance!

For that reason, I scribed a shim that I attached to the wall behind the vanity panel then screwed a board into that shim. This effectively clamps the table to the wall.  Thing is, the MFT table still moved a bit sitting on the now anchored table.

Well, I scribed another shim to go behind the MFT and made a similar clamp, but this time with star nuts to make clamping and unclamping the table completely trivial.  The shim also filled a gap behind the table and my "straight" wall; perfect for Sharpies.

Disappearing Miter Saw Stand

No, no, no, I haven't been robbed.  But I used to have my miter saw on a mobile stand and would wheel it away into the confines of my third-car bay.  Then the shop grew and, you know, free space was occupied.  It became a chore to get the saw out.  I sold the stand to the cheapest contractor on the face of the Earth (and whose gigantic pickup leaked oil on my driveway, thank you...) and made a stand for it that is dual use.

The initial idea was motivated by a mobile stand presented in Small-Shop Solutions.  Rather than flip to move one of two tools into operating position, mine flips the saw up or a work table up.

EDIT: Since I posted this, I've had a few questions about it.  Rather than add many more photos to this entry, I decided to run a short video demonstrating the table and going over the various build details.   The rest of this entry still has the original photos and text, but if you want video, here it is:

Voilà, the table in "work table" position.  The table top is coplanar with my MFT-1080 just to the left.  Underneath, besides various accessories I don't know where to store, is the miter saw.

Here you can see how the table top is coplanar with the MFT-1080 (on my stand; more in another posting later).

The table's position is locked with a cotterless pin on both sides of the rotating top.

Here's an action shot of rotating the saw into position.  All that's needed is to pull the pins slightly and the table, by the weight imbalance, will pivot up slightly.  Just grab the saw by the carrying handle to complete the move.

The resulting work position.  Note how the saw sits lower than the work space that was on the other side.  This is due to the offset position of the pivot point on the tabletop.  The result is that the deck of the saw is coplanar to the MFT-1080 allowing me to use that table to hold stock to the left.

The base itself is on a Jet mobile base since occasionally I get really long stock that needs to be cut in the driveway.  An enhancement I plan to make soon is to use feather keys to lock a spacer into the side of the MFT-1080 and a cam-lock on the miter stand to fix the position of the saw relative to the MFT-1080.  With that in place, I can clamp stop blocks trivially to the MFT.

I'm soon planning on getting a couple MDF sheets drilled in the same pattern as the MFT top, including the MDF under the saw.  With that done, I can clamp boards to the deck by using the clamps used for the MFT rather than trying to clamp to the corrugated underside of the saw deck.

I get asked often enough about how to compute where the pivot point is on the swinging table and how to compute its thickness.  Not very difficult so here it is:

Build the stand with the sides getting maybe a couple inches below the bench height. Plan on where you'll put the pivot point (mine's in the middle of the rail).

Measure from the middle of the pivot point to the bench surface.  Let's call that B so this sounds technical.

Measure the height of your saw deck... from the surface it sits on to the deck.  Call that M (as in miter saw).

The overall thickness of the pivoting box is then B+B-M; it would be B+B for two surfaces that happen to be equal to the bench height but we want to lower one by the height of the saw deck hence the -M.

Make the pivot hole in the swinging box 'B' away from the side that is just a work surface.

The higher the sides of the overall stand are, the smaller B becomes making the swinging box thinner.  I wanted this thickness because I wanted to be able to easily reach in for the star knobs used to bolt the saw to the table.

Tuesday, May 25, 2010

French Knot Walnut Inlay in Cork Floor

I took several pictures while doing this project because the Festool "system" surrounding the OF-1400 router was key to its success (yes, yes, you could make your own rail or do it by hand, but this was fast, easy, and, uhm, smart :)  I also posted this entry in a shorter form on as part of a contest I won't win.

Here, I have my new entryway :)  I took 2 sheets of 1/8" luan plywood and cross-laminated them together to form a substrate that was water resistant and able to be directly attached to my slab foundation.  I covered it in a nice cork in a herringbone pattern; in dry-fitting the layout, I found the herringbone pattern didn't telegraph the seams as much as others.  I then laid out a French knot pattern with masking tape that by chance was very nearly the same width as my extremely bowed scrap of Walnut :)  Center stage is the OF-1400 router and a guide rail.  Those two made this process ridiculously easy.

To route the recesses in the cork, I'm using a 1/4" down-spiral bit.  My process is to route away the outside edges of the recesses then go back and eliminate the part left behind with another bit.  I just don't want to buy another bit to avoid the 10-minute pass with another bit to finish it off.

To make sizing the recess dead-nuts accurate, I'm going to use the KM-1 by Bridge City Toolworks.  The KM-1 works by using the size of your cutter (a router bit in this case) and the size of the dado (recess here) to create 2 offset fences that you'll see as I go along.

Set the 'orange' jaw of the KM-1 to the width of the bit your using.

Next, set the grey jaws of the KM-1 to the width of the walnut we're inlaying.  Keep it close by :)
Okay, this picture has several important points.  I swapped the stock base with the table widener.  In this case, I don't need the added stability the widener gives the OF-1400, but rather the added thickness that lets me use the guide stops off the back of the rail so the router is sitting on the stock instead of on the rail (I generally don't like it sitting on the rail since I never get the compensating support foot to lock solidly).  Note how I do not have the micro-adjusting screw of the guide stops attached between the stops.

Position the bit so it will cut the inside edge of the recess (inside meaning closest to the guide rail).

 I do use the dust guard, but find that I had much better visibility if I left the front part of the guard open.  It didn't change the dust collection much.  Remember that this is a down-spiral bit so most of the waste was left in the wake of the recess.

Measure from the guide rail to the inside edge of the recess as you'll use this to properly align the guide rail to the recess for all future cuts after the initial setup is completed.  Note that using a triangular bench rule like this one is nice since it butts up to the guide rail eliminating some error.

At this point, the router is ready to route the inside part of the recess, but we want this to be as "faster, easier, smarter" as can be so we need stop blocks calibrated with the KM-1.

In this picture, the leftmost guide block is first locked in place.  The KM-1 is placed between the guide blocks in the "long" position and the rightmost guide block is moved to touch it and locked into position.  The rightmost guide block will never be moved again in this project.  You'll notice something to the left of the leftmost guide block (apologize that it wasn't in the previous pictures).  That is the microadjuster for the MFK-700 as it fits perfectly on the rods and locks.  I use it as a well-sized stop block.  Butt it up against the left guide block and lock it in place.  You won't move it again.  Note that the "microadjust" ability is not used; it just sits there like a dumb brick marking a stop location; I plan on ordering another by digging through the Festool EKAT parts system.

Next, flip the KM-1 to use the "short" fence.  Loosen the leftmost guide block and move the router over until the KM-1's short fence is between the guide blocks.  I have a stop collar from a drill-bit set to the right of the leftmost guide block (right behind the KM-1... it's small).  It doesn't fit as well as I'd like, but worked.  Tighten this stop collar to mark the short fence location.  You can return your KM-1 to its hand-made walnut keeper chest now; the collars will mark our offsets.

This is where your router bit should be located right now after the adjustments using the KM-1 (if not, verify you didn't budge the guide rail).  Also remember that in my case, the tape is nearly identical width to my walnut; YMMV.

So now you are ready to crank through the routing operations.  Note in the original picture that several recesses line up within a French knot and to an adjoining French knot.  This is by design.  Even if your tape isn't perfectly laid out, using the guide rail aligned across all co-linear segments will make the whole that much better looking.

Here's the steps:
  1. Place the guide rail the distance noted above from the inside edge of a recess to route.  Clamp the rail because patching this cork is not an option.
  2. Loosen the leftmost guide block (one between stop collars) and slide the router until the guide block abuts against either stop collar.
  3. Plunge and route; I usually scribbled with a red Sharpie at the ends of a segment so I could be forewarned when looking through the router window.  Stop early as it is easy to square the corner with a sharp chisel.
  4. Loosen the leftmost guide block again and slide the router until the guide block abuts against the other stop collar.
  5. Plunge and route.
  6. Go to step 1 until you have no more recesses.  This is a very fast cycle.
Once the recesses are done, chuck up a straight bit (1/2" in my case) and freehand between the "moats" you just made to clear the waste between the edges.  This won't be neat.

Occasionally the rail didn't make it the whole way across the board.  It missed by inches.  For that, I inserted the rail connector into the rail and put a small piece of home-sawed veneer under the connector before clamping it down (veneer simply takes up the space under the connector so it doesn't pop the rail up).

Next up, cutting the walnut to length and mitered corners.  This isn't difficult, just tedious.  One tip: when 7 segments go in first-try and the eighth is being, uhm, fussy, check for new posts on your favorite forum.  Really, you just need to walk for a second.

So the rest of this has less to do with the router, but finishes the story of the floor.

Here's the walnut dry-fitted into the recesses.  You want a very snug fit, but try not to fully seat anything while dry-fitting since pulling it out can shred edges of the cork (this is why an up-spiral bit isn't appropriate).

 Paint the recesses with contact cement and let it flash off (about 30 minutes).

Paint the back of each inlay piece as well.  Note that I labeled each piece on the back to know where it goes.  Put pieces in order into the recesses.  I found that gently pushing them in and aligning the neighbors was the best way before pressing down hard.  Once you press down hard, it's done.  You'll want a scrap of cork glued to a block so you can use it to mallet the inlay pieces afterward to ensure they are seated and glued.

I used a shoulder plane to flush the walnut to the cork.  I preferred the medium shoulder plane mostly because the blade was just marginally wider than the walnut.  This greatly reduced the chances of striking the cork.

Voilà, the finished inlay!

For context, it will be in the entryway at the bottom of a set of stairs I refaced almost 2 years ago.  Here are some photos.
Bottom step will have the riser attached after the entryway is installed over the tiles you see to the right in the photo.
The landing has a French knot, as well.  It was done in a much more difficult manner; I was more stupid back then.  Oh, the lopped off corner of the entryway inlay is where the door is for the entryway so the entryway and landing will look the same from the top of the steps.  I have since installed base molding.
The bullnose of the existing treads was cut off and replaced with triple-beaded walnut.  Risers are home resawn lacewood veneer.  The corner molding you see here was custom cut and has continuous grain from the topmost step to the bottom.  Now, that molding is trivial, back then, it was tricky.  Guess I learned something :)

Galootish Tendancies

So I get labeled a Normite since I have a couple powertools and will dovetail drawers with a router jig before whipping out my favorite Dozuki to do them by hand.

Behold, photographic proof that I know how a hand plane works.

This 30"x30" panel of QS ribbon mahogany was jointed flat by hand on the underside.  The top surface was hand processed with the exception of the progressive tapered cove cut in by table saw using a jig I came up with and a program to print out the math involved.

The 2" wide band on the side of the cove is a flat area and the sides bevel down to 1/4"; since the inside edge of the bevel is along a tapered cove, it is tapered adding to the fun. This is a picture after I finished and swept the board.  Keeping the bevel even and the flats coplanar was half the fun of doing this by hand; at the widest, the cove is wider than my plane.

Here's the other angle showing the hamster bedding everywhere.  Note the ribbon curl.  Normal planing techniques tear-out like nobody's business.  I'm happy to report absolutely zero tear-out; not even a little tiny corner that needed sanding.
Da bucket... there were more on the floor.  Yeah, I know this isn't dust like my table saw generates, but a DC takes care of that.  This you have to sweep.  SWEEP!! Oh, my...

Here's the resulting board with the hand-shaped handles, done with a spokeshave and occasional scraping.

The result after 3 highly diluted coats of Seal-A-Cell rubbed in (handles aren't yet attached).  Glass smooth.  Drawer fronts will be cut from this and applied to the drawers I hope to glue-up tomorrow.

This whole drawer front stack was an original design and so far, I'm pretty happy with it.

Now, gimme back my damn power cord...

Simple Ways to Reduce Snipe with Lunchbox Planers

I wrote this a long time ago and, as the article states, my lunchbox planer was installed in a machine stand at the time (the trick relies on the planer being portable).  Although I describe everything here, photos are lacking.  I've since rolled a video explaining it all here: How to Eliminate Snipe on Lunchbox Planers.

Snipe sucks.  Even starts with an "S", too.

A lot of people have tips to gently lift the piece as it enters the planer and press down as it exits or the usual "you just need to make sure the tables are coplanar" (duh).  Thing is, the lift can introduce a hump in place of the snipe and even if the tables are coplanar, most are rather flimsy and flex under the variable load of the board (variable due to the change in loading as it enters the lunchbox).

Before getting a battleship (er, sorry, PM20 planer), I had a Ridgid 12" lunchbox and honestly I highly recommend it.  To avoid the problem with snipe, I wanted more stable tables on both sides especially for the long boards I was running through (my first projects had dimensions from 4'-8' routinely... don't ask).

My solution was to build two torsion boxes and top them with strips of melamine-coated particle (MCP) board.

I made them as torsion boxes on an assembly table to ensure they were flat.  Using strips of MCP allows the stock to move smoothly (with half the friction since half the box isn't covered) and allows a place for ejected chips to go rather than lodge under the stock.

Now, I have since put my planer inside a stand so removing it for a picture is an hour affair, but the way I used these tables was to put the planer on the assembly table shown with a table on the infeed side and the other on the outfeed side.  The stack of grey wood are shims to get the tables to the correct height.  Since the assembly table is built to be flat, I can place the planer anywhere on it with the in/out-feed tables and have everything work well.  This was extraordinarily useful with longer stock.

Today, I use this planer for drawer stock or small box stock (think small thin stuff).  Since I cannot put the planer on the assembly table, I put one of the auxiliary tables inside the planer.  Written on the side is 82mm, the thickness of these tables.  I always plane in metric since fractional millimeters are easier to deal with than imperial measures.

In my case, I'm reusing the tables I already made that were about 4' long, sized conveniently to fit on my assembly table in the original orientation.  If you like this idea and want to use it, I'd suggest making a 5' box so you have at least 2' on each side (maybe even longer).  Other than applying the offset, everything is the same as before, except no lifting or feeling shame that apparently your tables aren't as coplanar as the other guy (-psst!- he likely read it on a forum :)

Sunday, May 23, 2010

Adding "Feet" to a Flat-Bottomed Cabinet

So, I made a bench for planing, vice work, "SysPort", and clamp storage.  The sides and front were straight rails that sat on the ground.  Okay, okay, dumb idea given the "flatness" of the average garage.

Anyhow, I quickly and easily retrofitted "feet" into this bench (I didn't want to change the height as it was matched to an MFT/3).
This is how I did it: first, blue-tape the end of each foot on each side of a rail. Next, use a bearing-piloted rabbeting bit (say that 3 times fast...) to make the pattern in place.  That is, use the bearing against the rail itself and the top thickness of the rail will get the rabbet.  The rabbet is between the feet you delineated with the blue tape and naturally a stopped rabbet like this will have nicely rounded ends.
Now, to finish.  For this, I used a bearing-piloted pattern bit (okay, that's redundant).  The pattern was the rabbeted recess I just created and it will flush away the rest of the rail to complete creating the feet.
Voilà, the result: